Inspiration
Sparked by the need to simplify document processing for students, researchers, and professionals overwhelmed by lengthy PDFs, reports, and texts. Envisioned a tool that harnesses AI to deliver concise summaries, supports multiple languages, and provides a seamless, browser-based experience—no complex software installations. Goal: Empower users to save time and focus on insights, not information overload.
What it does
Transforms how users interact with documents via a web-based platform. Features: Upload Files: Drag and drop PDF, TXT, or DOCX files for instant processing. Generate AI Summaries: Extracts key insights, condensing documents to ~20% of their original length using advanced AI (powered by Grok AI). Multilingual Support: Summarizes and translates content into over 20 languages, such as English, Spanish, French, and Chinese. Export Results: Downloads summaries as text files, with planned support for polished PDF/PPT presentations. Privacy-First: Processes files in-browser or via secure serverless functions, ensuring data stays safe.
Intuitive, no registration required, optimized for speed and accessibility.
How we built it
Developed with a modern tech stack: Frontend:
HTML5 and JavaScript for a responsive UI with file uploads, language selection, and summary display. Tailwind CSS for a sleek interface with animations (e.g., typewriter effect, progress bar).
...ackend:
Vercel serverless function (/api/upload.js) using formidable to parse uploaded files in memory, ensuring compatibility with Vercel’s stateless environment.
AI Simulation: Used sample content in multiple languages to mimic AI summarization; planning Grok API integration for production.
Deployment: Hosted on Vercel for scalability.
Validation: Client-side checks for file type (PDF, TXT, DOCX) and size (<4.5 MB).
Challenges we ran into
Vercel Deployment Issues: File uploads worked on localhost but failed on Vercel due to its stateless environment. Resolved by processing files in memory with formidable and disabling Vercel’s default body parser.
Cross-Browser Compatibility: Ensured drag-and-drop and file input functionality worked across browsers through rigorous testing.
AI Integration: Simulating AI summarization without a live API was challenging; used sample content as a placeholder, limiting real PDF processing.
Accomplishments that we're proud of
Seamless UI/UX: Crafted a visually stunning interface with Tailwind CSS, featuring gradients, blur effects, and engaging animations. Vercel Compatibility: Adapted the app for Vercel’s serverless environment with a robust /api/upload.js function. Multilingual Support: Implemented a language selector with sample content in 7 languages, laying the foundation for global accessibility. Scalable Design: Built a modular codebase ready for future integrations like real AI summarization and cloud storage.
What we learned
Serverless Architecture: Gained expertise in Vercel’s serverless functions, handling file uploads without disk access, and managing payload limits. Frontend Optimization: Balanced aesthetics (animations, gradients) with performance for a smooth user experience. File Processing: Mastered parsing multipart form data in a serverless environment using formidable. User-Centric Design: Emphasized clear error messages (e.g., file size/type validation) and progress feedback for usability. Deployment Debugging: Learned to analyze Vercel logs and browser Network tab errors for troubleshooting.
What's next for ExamCao
Real AI Integration: Connect to Grok API or a library like pdf.js to extract and summarize text from actual PDFs, replacing sample content. Cloud Storage: Integrate AWS S3 or Cloudinary to handle larger files, bypassing Vercel’s 4.5 MB limit with presigned URLs. Analytics Dashboard: Introduce user analytics (e.g., number of files processed, time saved) to showcase value, as reflected in the stats section.
Log in or sign up for Devpost to join the conversation.